Skip to page content
SAVRN
Search Plan a project

Dataset

dpo-german-interpreted

by Jonas Lindh jonasaise/dpo-german-interpreted

This dataset contains Direct Preference Optimization (DPO) pairs translated and culturally adapted into German from an English source dataset (e.g. ultrafeedback or safe-rlhf).

Rows—
Configurations—
Size27.6 MB
Licensemit
AccessPublicly accessible
Monthly Downloads—

Dataset Card

By Jonas Lindh, published under mit, revision 5b541a64affd.

Dataset Card for DPO German Interpreted

Dataset Description

This dataset contains Direct Preference Optimization (DPO) pairs translated and culturally adapted into German from an English source dataset (e.g. ultrafeedback or safe-rlhf).

Origin and Methodology

  • Model Used for Synthesis: kimi-k3 (via local endpoint)
  • Translation Strategy: The dataset was generated using a specialized prompt designed to prevent "translationese" and direct literal translation. The kimi-k3 model was instructed to heavily localize the prompt and the chosen response to fit German culture and idioms. Furthermore, it intentionally generated a rejected response that suffers from poor translation, excessive English loan words, or US-centric hallucinations. This provides a high-quality, targeted signal for DPO/RLHF to teach models to favor authentic, culturally grounded German over robotic dictionary translations.

Structure

The dataset contains the following fields: - prompt: The culturally adapted German prompt. - chosen: The high-quality, authentic German response. - rejected: The poorly translated, uncalibrated response. - target_language: The target language of the row. - dataset_source: The source dataset name. - is_safety_focused: Boolean indicating if this is a safety alignment sample.

License

This dataset inherits the license from its source dataset (MIT or Apache 2.0).

Details

Repository
jonasaise/dpo-german-interpreted
Publisher
Jonas Lindh
Task category
Not stated by the source
Tags
dpo, rlhf, ultrafeedback
Size category
Not stated by the source
Languages
ge
Revision
5b541a64affdfdbaeeaa1548539f25780f9da788
Last updated
2026-09-25

Files

183 files, 27.6 MB in total.

Data181 files · 27.6 MB
Documentation1 file · 1.6 KB
Repository1 file · 2.5 KB
Every file
FileTypeSizeSHA-256
data/train-1790246652.parquetData143.3 KB495e8993fed3
data/train-1790247019.parquetData171.4 KBc5f2c93fb87d
data/train-1790247386.parquetData148.7 KB2a034b9940e4
data/train-1790247747.parquetData141.0 KBc77256811c83
data/train-1790248136.parquetData149.1 KB7f16f206152b
data/train-1790248523.parquetData148.7 KBb69dc0376e02
data/train-1790248815.parquetData152.9 KB0b93010f6db3
data/train-1790249188.parquetData150.6 KB4dd3237dd316
data/train-1790249520.parquetData135.8 KB8ca155d3e978
data/train-1790249863.parquetData159.1 KB1f4b3961febf
data/train-1790250203.parquetData159.9 KBbc38c8b5275c
data/train-1790250562.parquetData147.3 KB845bebee9408
data/train-1790250908.parquetData161.5 KB47f1f7135800
data/train-1790251281.parquetData143.3 KBb77456dc6792
data/train-1790251661.parquetData154.8 KB9182a3b9eb38
data/train-1790252000.parquetData143.6 KBfdd1161daa2a
data/train-1790252341.parquetData164.1 KB35f8bfb88126
data/train-1790252635.parquetData134.7 KBf68154d5145e
data/train-1790252946.parquetData151.1 KB85a35d959bef
data/train-1790253308.parquetData159.1 KBc332e4ae4dc7
data/train-1790253705.parquetData167.8 KBb97ab6680ecd
data/train-1790254054.parquetData137.3 KB451b904e1b1a
data/train-1790254404.parquetData131.3 KBb2e0ba463649
data/train-1790254721.parquetData139.3 KB5844cfa255a2
data/train-1790255039.parquetData144.2 KBe31f57bb4110
data/train-1790255436.parquetData156.2 KBcd18d025049e
data/train-1790255778.parquetData135.1 KBce8178b6c203
data/train-1790256151.parquetData145.4 KBbd0d5df5af48
data/train-1790256505.parquetData170.3 KB88aa1c2a1563
data/train-1790256844.parquetData143.9 KB77c5beabf56f
data/train-1790257219.parquetData158.6 KB4163077ebc2c
data/train-1790257599.parquetData132.0 KB537f40f3cbf1
data/train-1790257996.parquetData153.5 KB619b245854cc
data/train-1790258293.parquetData126.6 KB1f578921107f
data/train-1790258625.parquetData156.9 KB33a7e6fdb86c
data/train-1790259005.parquetData155.5 KBbc01012d8d6a
data/train-1790259365.parquetData163.7 KB3c4fae6dcc0a
data/train-1790259680.parquetData155.5 KBd96f1cf87185
data/train-1790260024.parquetData152.9 KB2c159c6f3fdc
data/train-1790260375.parquetData137.5 KB3519454abacf
data/train-1790260700.parquetData162.1 KB4bf38e53bab1
data/train-1790261065.parquetData146.8 KBb77f2c80b7fa
data/train-1790261422.parquetData164.8 KB0c786f45dacf
data/train-1790261688.parquetData132.5 KBf16cc65dfe81
data/train-1790262008.parquetData156.4 KBc60d84060fb4
data/train-1790262366.parquetData164.7 KBcc143a9c33c2
data/train-1790262729.parquetData167.1 KB0f339433b3df
data/train-1790263056.parquetData153.2 KB326dc12a7269
data/train-1790263396.parquetData155.4 KB42aef71fd997
data/train-1790263752.parquetData149.5 KB373c416b0d0d
data/train-1790264057.parquetData160.2 KBec2e25c33d5f
data/train-1790264400.parquetData157.2 KB478fec4ac6be
data/train-1790264718.parquetData149.9 KB2420988336a9
data/train-1790265056.parquetData144.1 KB1ff4926a2b5f
data/train-1790265407.parquetData151.4 KB553a8fd647ad
data/train-1790265769.parquetData166.8 KB62255ed0596f
data/train-1790266105.parquetData157.3 KB7415061a5cd1
data/train-1790266450.parquetData170.9 KB20e97e1b53ee
data/train-1790266823.parquetData159.3 KB2aa6127768a9
data/train-1790267134.parquetData151.1 KB9903d26e5f10
data/train-1790267460.parquetData158.5 KB267589c3a68e
data/train-1790267850.parquetData169.6 KBd9c4a890de27
data/train-1790268162.parquetData148.5 KB72b6cb7c88a5
data/train-1790268475.parquetData145.4 KB844fbce67f8b
data/train-1790268880.parquetData159.3 KB1e0c88b30daa
data/train-1790269179.parquetData140.2 KBd4683c241430
data/train-1790269505.parquetData166.4 KB0193680778ac
data/train-1790269859.parquetData158.0 KBe0da557eb8dc
data/train-1790270230.parquetData146.9 KB6e190ff54375
data/train-1790270600.parquetData156.2 KBe184cc57d097
data/train-1790270949.parquetData160.7 KBea1b3296de4f
data/train-1790271326.parquetData155.4 KB4fa143fe20db
data/train-1790271672.parquetData143.7 KBb938e7088a28
data/train-1790271999.parquetData180.3 KB09ee6c038902
data/train-1790272279.parquetData124.0 KBc48c9a8f50ff
data/train-1790272597.parquetData140.2 KB2d3457600003
data/train-1790272898.parquetData158.3 KBd19a81fb9af0
data/train-1790273313.parquetData159.3 KB6afc0706fe42
data/train-1790273730.parquetData180.9 KBcb1ac19e9ae4
data/train-1790274093.parquetData165.3 KB11b0322958cb
data/train-1790274429.parquetData150.2 KBe01eefe6e15f
data/train-1790274719.parquetData147.4 KBf874206eaa2a
data/train-1790275038.parquetData154.4 KBabd1a1af9c03
data/train-1790275398.parquetData149.1 KBaa1e96eccf66
data/train-1790275721.parquetData152.1 KBedfe66078de8
data/train-1790276016.parquetData150.3 KBa082379c22c5
data/train-1790276394.parquetData161.1 KB157c98f17eb4
data/train-1790276725.parquetData153.9 KBce2f8dcf6ff5
data/train-1790277053.parquetData141.8 KBb5392ce930a7
data/train-1790277387.parquetData165.6 KBe481a513e78f
data/train-1790277731.parquetData158.4 KB000304a63cc6
data/train-1790278054.parquetData144.0 KB190c7f79090b
data/train-1790278401.parquetData161.9 KB16a329a80b6e
data/train-1790278778.parquetData150.7 KBe7b75ce18293
data/train-1790279218.parquetData158.4 KBe81f2a42a6c7
data/train-1790279599.parquetData152.2 KB01303b72d317
data/train-1790279951.parquetData147.3 KB74dbfeba6736
data/train-1790280318.parquetData161.4 KB926162328702
data/train-1790280655.parquetData135.1 KBc0c4e1b8fcd5
data/train-1790281017.parquetData160.6 KBf2f78684b032
data/train-1790281298.parquetData138.6 KBe0a98ba90647
data/train-1790281706.parquetData166.6 KB246cddd6e57c
data/train-1790282061.parquetData154.4 KB1aded597f281
data/train-1790282393.parquetData155.6 KBc96b42ce34a8
data/train-1790282707.parquetData143.1 KB18d610b5fa34
data/train-1790283035.parquetData160.5 KBa61bea26343d
data/train-1790283369.parquetData149.1 KBf330785fd990
data/train-1790283684.parquetData147.4 KB5d000507b159
data/train-1790284022.parquetData167.1 KB1870e7fc386f
data/train-1790284328.parquetData153.1 KBa222336bdb05
data/train-1790284673.parquetData142.9 KB2d6b53e98dc1
data/train-1790284993.parquetData148.0 KBdcf591203d10
data/train-1790285357.parquetData157.8 KB32cd4c076e98
data/train-1790285650.parquetData150.3 KB9a62fde62a83
data/train-1790285991.parquetData180.4 KBea58922a3865
data/train-1790286306.parquetData139.1 KB3c3e28278997
data/train-1790286644.parquetData141.0 KB9a88bd830006
data/train-1790286997.parquetData147.9 KB91482b585262
data/train-1790287325.parquetData146.9 KB03a64748bfc7
data/train-1790287697.parquetData155.3 KB0adb960decb9
data/train-1790288024.parquetData152.7 KB7d0e77d7ed50
data/train-1790288383.parquetData154.6 KBe0c2c8d7f507
data/train-1790288729.parquetData165.7 KB285bfeef9c4b
data/train-1790289055.parquetData154.3 KB0879d346206c
data/train-1790289316.parquetData136.0 KBec10e67bfe5a
data/train-1790289621.parquetData127.3 KB98caf69132dc
data/train-1790289949.parquetData156.2 KBa75e69c47083
data/train-1790290302.parquetData165.3 KB00ed35910153
data/train-1790290674.parquetData169.8 KBdf5273da9dec
data/train-1790291026.parquetData162.1 KB23befbfbe9a4
data/train-1790291329.parquetData149.4 KB3776dc7a5086
data/train-1790291662.parquetData166.0 KB7e8d0ca12059
data/train-1790291977.parquetData146.2 KBfc13cecc1b93
data/train-1790292287.parquetData155.7 KB996f3f7e5180
data/train-1790292619.parquetData148.5 KBf2de9cba00c4
data/train-1790292984.parquetData168.9 KB46b3e01ca365
data/train-1790293257.parquetData134.0 KB71ba74690b6a
data/train-1790293574.parquetData150.4 KB7c8e6b038e5e
data/train-1790293910.parquetData140.5 KBbda648836903
data/train-1790294181.parquetData125.8 KBbe74d13f33e5
data/train-1790294509.parquetData168.2 KB6f56c4f29047
data/train-1790294873.parquetData161.2 KB096882bb6276
data/train-1790295175.parquetData155.2 KB34e2ec530dc7
data/train-1790295533.parquetData166.3 KBe14a0f8449bf
data/train-1790295870.parquetData164.5 KBe90381d3efdb
data/train-1790296163.parquetData154.7 KB305af2d6f4b8
data/train-1790296421.parquetData128.1 KBb73430b369d9
data/train-1790296796.parquetData159.5 KB5f684cc847c5
data/train-1790297195.parquetData186.6 KB467f1dfb1dd6
data/train-1790297517.parquetData149.3 KB39b5cbaaa74f
data/train-1790297804.parquetData139.9 KB7de2d30a2a5f
data/train-1790298110.parquetData156.9 KBec0a478b724a
data/train-1790298457.parquetData153.3 KB6c98ba5b5da8
data/train-1790298771.parquetData151.5 KB5307baf580a4
data/train-1790299082.parquetData153.7 KB485684bb5dbd
data/train-1790299405.parquetData125.0 KB07caf60b2a70
data/train-1790299738.parquetData160.3 KB752ebfe9d213
data/train-1790300071.parquetData147.3 KBd5e3527cdc15
data/train-1790300361.parquetData142.8 KBcf62da30fc6c
data/train-1790300723.parquetData165.3 KB182114001f97
data/train-1790301026.parquetData135.4 KBe2d67bdc8796
data/train-1790301354.parquetData148.9 KB97d90263347c
data/train-1790301666.parquetData150.2 KB906e5cc2d680
data/train-1790302041.parquetData167.0 KB1a6dae136502
data/train-1790302364.parquetData151.1 KBa9d98b1aa3b7
data/train-1790302765.parquetData162.2 KB63f22c505398
data/train-1790303135.parquetData162.2 KBd257e55e8de7
data/train-1790303461.parquetData149.5 KBbd27e76eac11
data/train-1790303875.parquetData186.3 KB0d50eb483414
data/train-1790304186.parquetData138.7 KBc6c628b8c62b
data/train-1790304523.parquetData152.2 KB74d29a4fa776
data/train-1790304851.parquetData150.9 KBb5255c610eb8
data/train-1790305212.parquetData168.4 KB1514270603e6
data/train-1790305548.parquetData142.1 KBaca899b3db5d
data/train-1790305877.parquetData145.0 KB03fb72992bcd
data/train-1790306221.parquetData157.5 KB5d44d5fff002
data/train-1790306576.parquetData162.9 KBfcd7496f7bbf
data/train-1790306907.parquetData134.1 KB5de68cc4c714
data/train-1790307221.parquetData156.8 KB92a3ec779290
data/train-1790307540.parquetData150.3 KBa0af2625bd9e
data/train-1790307839.parquetData123.2 KB69ed412b05f2
README.mdDocumentation1.6 KB—
.gitattributesRepository2.5 KB—

License and Download

License
mit
Access
No access gate
Download from Jonas Lindh

Released by Jonas Lindh through its official repository on Hugging Face. Read the license.